[Dev] Re: refcounted

[John Anderson]

In place of the pinning code I put an assert where I removed SetPinned 
so we can find them, but I think it's very unlikely we'll ever go back 
to pinning. This caused the ItemCollectionUnit test to fail, so I had to 
back them out.

Morgen's idea is great and I already made a pass on OnInit. However, I 
think there may be a variety of things we need to initialize besides the 
repository and I need to figure out what the smallest set of things to 
initialize -- so we don't have zillions of separate initialization 
routines. I haven't been able to finish that yet.

Also, I was curious to know why you prefer to leave non-pinning default 
when we have no plans to use it and Ted, Morgen, Lisa and I prefer the 
opposite?

John

Andi Vajda wrote:

>
>> Removing all the pinning code is at your option as well.
>
>
> It would be prudent to leave a comment in place of the pinning code 
> locations in case we need to revert to pinning if the new 'refcounted' 
> mode doesn't pan out.
>
> Andi..
>
>>
>> Andi..
>>
>> On Thu, 28 Oct 2004, John Anderson wrote:
>>
>>> Hi Andi:
>>>
>>> Let me know if you plan to default to refcounted -- if so I'll be 
>>> able to remove the remaining SetPinnned code in Chandler without 
>>> breaking the unit tests.
>>>
>>> John
>>>
>>> Ted Leung wrote:
>>>
>>>> +1
>>>>
>>>> On Oct 25, 2004, at 1:35 PM, Morgen Sagen wrote:
>>>>
>>>>> I also agree that we should just have the repository default to 
>>>>> the new refcounted mode.
>>>>>
>>>>> On Oct 25, 2004, at 6:59 AM, Lisa Dusseault wrote:
>>>>>
>>>>>> I'm tending to agree with John here.  What's the cost to keeping 
>>>>>> the refcounted and non-refcounted modes?  Are we going to test 
>>>>>> both modes? apply fixes to both modes?  document, etc?  Unused 
>>>>>> code goes stale quickly.
>>>>>>
>>>>>> Lisa
>>>>>>
>>>>>> On Oct 24, 2004, at 6:36 PM, John Anderson wrote:
>>>>>>
>>>>>>> I already did that -- that wasn't my point. I was suggesting 
>>>>>>> removing the refcounted kwd since I can't think of any reason 
>>>>>>> you'd want to use the repository non ref counted -- and if you 
>>>>>>> can think of a reason, the kwd should be nonrefcounded so no kwd 
>>>>>>> is ref counted (i.e. a sensable default)
>>>>>>>
>>>>>>> Andi Vajda wrote:
>>>>>>>
>>>>>>>>
>>>>>>>> You're welcome to turn it on in application/Application.py, 
>>>>>>>> line 207.
>>>>>>>> Same applies to the -exclusive option, line 206, the app needs 
>>>>>>>> to decide what it wants to do there as well.
>>>>>>>>
>>>>>>>> Andi..
>>>>>>>>
>>>>>>>> On Sun, 24 Oct 2004, John Anderson wrote:
>>>>>>>>
>>>>>>>>> Andi:
>>>>>>>>>
>>>>>>>>> Now that we have the refcounted option in the repository, I'd 
>>>>>>>>> vote for making it the default and eliminating the option of 
>>>>>>>>> non-refcounting altogether -- unless you can think of any 
>>>>>>>>> advantage of non-refcounting.
>>>>>>>>>
>>>>>>>>> John
>>>>>>>>>
>>>>>>> _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
>>>>>>>
>>>>>>> Open Source Applications Foundation "Dev" mailing list
>>>>>>> http://lists.osafoundation.org/mailman/listinfo/dev
>>>>>>
>>>>>>
>>>>>>
>>>>>> _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
>>>>>>
>>>>>> Open Source Applications Foundation "Dev" mailing list
>>>>>> http://lists.osafoundation.org/mailman/listinfo/dev
>>>>>
>>>>>
>>>>>
>>>>> _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
>>>>>
>>>>> Open Source Applications Foundation "Dev" mailing list
>>>>> http://lists.osafoundation.org/mailman/listinfo/dev
>>>>>
>>>> ----
>>>> Ted Leung                 Open Source Applications Foundation (OSAF)
>>>> PGP Fingerprint: 1003 7870 251F FA71 A59A  CEE3 BEBA 2B87 F5FC 4B42
>>>>
>>>> _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
>>>>
>>>> Open Source Applications Foundation "Dev" mailing list
>>>> http://lists.osafoundation.org/mailman/listinfo/dev
>>>
>>>
>>>
>> _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
>>
>> Open Source Applications Foundation "Dev" mailing list
>> http://lists.osafoundation.org/mailman/listinfo/dev
>>